|
Activity Number:
|
37
- Combining Data and Use of Administrative Lists
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, July 30, 2017 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Survey Research Methods Section
|
|
Abstract #323748
|
|
|
Title:
|
Statistical Analysis with Linked Data
|
|
Author(s):
|
Ying Han* and Partha Lahiri
|
|
Companies:
|
University of Maryland, College Park and University of Maryland, College Park
|
|
Keywords:
|
Record Linkage ;
Generalized Linear Regression ;
Linkage Errors ;
Measurement Error ;
Bias Correction ;
File Matching
|
|
Abstract:
|
Probabilistic Record Linkage has made it possible to combine multiple data sets from different when a unique data set with all necessary information is not available or creating a new set is time consuming and extremely costly. Linkage errors are inevitable in the linked data sets because of the unavailability of error-free unique identifiers and because of possible errors in measuring or recording. Small linkage errors can lead to substantial bias and increased variability when estimating the relationship among the variables. Recent research has focused more on bias correction for linear regression analysis and assumes the linkage process is complete, i.e. all records on the two data sets are linked. In this paper, we extend these ideas to allow generalized linear regression analysis in the presence of linkage errors and incomplete linkage. We consider a special case where survey data and administrative data are linked. A generalized linear model of linked data with linkage errors is proposed. Simulation results will be presented to evaluate the performance of the proposed estimators, which account for linkage errors.
|
Authors who are presenting talks have a * after their name.
